Discussion

In the literature, the most common LHON mutation is at the m.11778 locus on MT-ND4.^{Reference Mackey, Oostra and Rosenberg5} However, in the genetics review presented here, it accounted for only 44.4% of classic mutations and 31.7% of all mutations. The most common mutation was at the m.14484 locus on MT-ND6, accounting for 48.9% of classic mutations and 34.9% of all mutations (Table 1). This may be accounted for by the known founder effect of French Canadian families with the m.14484 mutation.^{Reference Laberge, Jomphe and Houde10} In the small chart review, the m.11778 mutation accounted for the majority (50%) of mutations. This mutation is also generally associated with the worst visual outcome.^{Reference Newman, Carelli, Taiel and Yu-Wai-Man1} Four of the 7 patients in the chart review with this mutation ended up with vision of counting fingers (CF) or worse in at least one eye (Table 2). The mutation at the m.14484 locus has been shown to have the best visual prognosis, showing improvement in up to 50% of patients.^{Reference Yu-Wai-Man, Vortuba, Moore and Chinnery6} There were two symptomatic patients (Patients 9 and 11) in our series who had a mutation at this locus, and both demonstrated visual deterioration over time rather than improvement (Table 2).

Overall, female representation and older age at presentation were more common in this series than previously reported in the literature. The male-to-female ratio of LHON mutations in our genetics review was 1:1 and in the chart review of symptomatic patients was 1.4:1. This unexpected distribution may be due to the small sample size. However, there is growing evidence to suggest that the incidence of LHON in women is higher than previously thought. A large international study looking at the demographics of LHON determined a male-to-female ratio of 3:1,^{Reference Poincenot, Pearson and Karanjia11} lower than the previously reported 5:1 ratio. They also found that below the age of 5 or above the age of 45 at time of onset, the sex ratio was 1:1.^{Reference Poincenot, Pearson and Karanjia11} A significant proportion (40%) of the patients in our genetics review were 40 or older at the time of genetic testing, and this may explain the higher female ratio in our study. However, given that the genetics review could only provide age at time of testing, it is not a true representation of age at symptom onset. Our chart review found that only two of the 12 symptomatic patients had symptom onset after the age of 35. The 2020 large international study found that 10% of patients developed symptoms after age 50.^{Reference Poincenot, Pearson and Karanjia11}

The vast majority of mutations implicated in LHON are found in mitochondrial DNA on the genes coding complex I subunits of the respiratory chain (NADH-ubiquinone oxidoreductase).^{Reference Catarino, von Livonius and Priglinger12} Mutations causing amino acid substitution lead to dysfunction in the respiratory complex I, whose function is to process oxygen and create energy within the cell. Disruption of this process translates to subacute loss of retinal ganglion cells and their axons, which comprise the optic nerve.^{Reference Kirches13} While many of the novel mutations proposed here (Table 1) are not previously documented to cause LHON, they do all cause an amino acid substitution within the respiratory complex I, and therefore have a plausible mechanism for causing disease. Given the lack of clinical information available on most of the 63 patients, it cannot be ascertained if they all had optic neuropathies or whether other symptomology or family history may have prompted genetic testing. There are multiple known mutations found on the MT-ND1, -ND4, and -ND6 genes that are associated with other conditions such as Leigh syndrome, MELAS, and Parkinson’s disease.^{Reference Manickam, Michael and Ramasamy14} For this reason, all possible novel mutations included in the genetics review were searched in the literature to determine any previously identified associations. Multiple novel mutations in our genetics review were found in the literature in cases of LHON, supporting that these mutations are likely pathogenic. These include the m.4024A>T,^{Reference Korkiamäki, Kervinen, Karjalainen, Majamaa, Uusimaa and Remes15} m.3866T>C,^{Reference Zhou, Qian and Zhang16,Reference Ji, Zhang and Lu17} and m.14459G>A^{Reference Tarnopolsky, Baker, Myint, Maxner, Robitaille and Robinson18–Reference Jun, Brown and Wallace20} mutations (Table 1). The only mutation that yielded an association outside of LHON was m.11240C>T, which has been linked to Leigh syndrome in one previous case.^{Reference Xu, Li and Du21} All patients in the small chart review with novel mutations demonstrated optic neuropathies and homoplasmy of their mutation (Table 1), which is consistent with LHON.^{Reference Yu-Wai-Man, Vortuba, Moore and Chinnery6} These mutations were m.4231A>G, m.3367A>G, and m.14208T>C (Table 2). None of these have previously been documented to be of clinical significance in the literature.

The ophthalmological findings of patients in the chart review are consistent with what has previously been reported in the literature. Most patients demonstrated optic nerve pallor, primarily in the temporal quadrant. Two patients had hyperemic nerves, which can be seen in the early stages of disease.^{Reference Shemesh, Sood and Margolin22} Patients who underwent OCT of the optic nerves demonstrated corresponding thinning or thickening of the nerves, depending on the timing of testing. On MRI, patients showed a range of normal optic nerves, hyperintense signal, edema, and thinning; all features which have been previously reported in cases of LHON.^{Reference Blanc, Heran, Habas, Bejot, Sahel and Vignal-Clermont23} Two patients demonstrated normal optic nerves, which can be seen at presentation, but eventually optic atrophy is expected to develop in all patients with symptomatic LHON.

Although LHON is generally nonsyndromic, various systemic findings have been reported in patients with LHON. Two female patients (Patients 4 and 7) in this review suffered from a concurrent diagnosis of MS, compatible with numerous reports and case series in the literature describing this co-phenomenon. Harding (1992) described 11 patients with LHON and MS,^{Reference Harding, Sweeney and Miller24} hypothesizing that the presence of an LHON mutation in the mitochondria can be an instigator for MS. This is most often seen in female patients, and a previous study found that 45% of females with the m.11778 mutation have an MS-like illness,^{Reference Riordan-Eva, Sanders, Govan, Sweeney, Da Costa and Harding25} as demonstrated in Patients 4 and 7. Female patients with LHON who have no MS-like symptoms are still more likely to have MRI abnormalities suggestive of MS.^{Reference Harding, Sweeney and Miller24}

Three other patients in the chart review suffered from serious medical illnesses. Patient 5 was small for gestational age with abnormal movements, gait difficulty, and incoordination representing a possible neuromuscular disease of unknown etiology. Although not formally diagnosed in this patient, this could be compatible with the entity described as “LHON-plus” in which LHON patients have systemic neurological involvement.^{Reference Nikoskelainen, Marttila and Huoponen26} Patient 12 had viral meningitis as an infant with secondary hemiplegia, as well as a fibrosarcoma removed at a young age. Patient 13 had coexistent diagnoses of lupus, rheumatoid arthritis, and polycystic ovarian syndrome. There is no known correlation in the literature between these diseases and LHON, and neither of these two patients had a definitive diagnosis of LHON. The remainder of the patients were either healthy or experienced milder nonspecific systemic symptoms.

Limitations of this study include the retrospective nature of the data collection and the small sample size of patients with full clinical information available for review. For patients found to have possible novel mutations, without a full chart review, it is impossible to know whether the mutation is clinically significant, an incidental finding, or related to a different disease. For this reason, the correlation between laboratory findings and clinical findings remains incomplete in regard to these possible novel mutations. Mutations of low homoplasmy in particular are less likely to represent LHON disease, but were included here for completeness. The incidence of 1 in 77,000 found in this study is lower than previously reported,^{Reference Carelli, Carbonelli and de Coo2} even when all possible MT-ND1, -ND4, and -ND6 mutations were included. Although LHSC is the reference laboratory for the health region, it is conceivable that patients had testing completed elsewhere, and would therefore be missed in our calculation of incidence. Alternatively, this could reflect a lower incidence of LHON in the Canadian population, a possibility which requires further investigation to confirm. Conversely, the LHSC laboratory also receives samples from outside its catchment area of Southwestern Ontario, and therefore these results may not be specifically representative of the region. This study did not analyze other genes more rarely implicated in LHON, such as mitochondrial cytochrome B and ATP6, due to limited availability of comprehensive chart data. Future studies with access to clinical data regarding all implicated mutations are essential for clearly outlining the clinical picture of LHON in the context of both classic mutations and possible novel mutations. If these novel mutations can be identified in other patients with an appropriate clinical picture, this would help confirm their role in LHON.

Currently, there is scant literature documenting LHON in Canada, and there are no previously reported studies describing LHON in Southwestern Ontario. The data from this study contribute to a growing body of literature on this rare disease, helping to further our understanding of its natural history. Overall, the data collected in this study are consistent with that previously reported about LHON, with a few demographic exceptions. The higher proportion of m.14484 mutations may be accounted for by a founder effect of French Canadians migrating across Ontario. Higher proportions of affected females and older patients may actually be reflective of the true LHON patient population, as suggested by other larger studies.^{Reference Poincenot, Pearson and Karanjia11} Finally, data from this study add to the literature on possible disease-causing mutations for LHON. Given our findings, clinicians should test for LHON mutations in patients outside of the historical demographic bracket and should take care to consider mutations reported to be of unknown significance as potential pathogenic mutations.